The Rebel AI team raised a record amount of cash after the machine learning split

Break away from the group artificial intelligence Researchers raised a record-breaking first round of financing for a new startup involving general artificial intelligence, which marked the latest attempt to create an organization to ensure the security of the most powerful technology of the era.

The organization parted ways with OpenAI, which was established in 2015 with the support of Elon Musk to ensure that the super-intelligent AI system will one day not lose control and harm its manufacturers. According to two people familiar with the matter, after receiving Microsoft’s $1 billion landmark investment in 2019, the group’s development direction has diverged.

The new company Anthorpic is led by Dario Amodei, one of the founders of OpenAI and the organization’s former head of artificial intelligence security.it Already proposed The first round of financing was US$124 million. According to research firm PitchBook, this is the most money raised for artificial intelligence teams trying to build universally applicable artificial intelligence technology, not teams that apply the technology to specific industries. According to the data disclosed in the company’s documents, this round of financing valued Anthorpic at US$845 million.

The investment was led by Jaan Tallinn, an Estonian computer scientist behind Skype. Former Google CEO Eric Schmidt and Facebook co-founder Dustin Moskowitz also supported the joint venture.

According to one estimate, OpenAI’s breakthrough began when Amodei resigned in December, and has since grown to include nearly 30 researchers.They include Daniela Amodei, the president of Amodei’s sister Anthorpic, and a group of researchers working on GPT-3 and OpenAI breakthrough Automatic language systems, including Jared Kaplan, Amanda Askell, Tom Henighan, Jack Clark and Sam McCandlish.

Two years ago, OpenAI changed direction when it sought support from Microsoft to satisfy its growing desire for computing resources to support its deep learning system. In return, it promised the software company the priority to commercialize its discovery.

Oren Etzioni, head of the Institute of Artificial Intelligence founded by the late Microsoft co-founder Paul Allen, said: “They were originally a non-profit organization that aimed to democratize artificial intelligence.” Obviously, when you get a billion dollars, you have to generate returns. I think their trajectory becomes more corporate.”

OpenAI attempts to isolate Microsoft’s research on artificial intelligence security from its new business operations by restricting Microsoft’s presence on its board of directors. However, according to a person familiar with leaving the organization, this still leads to internal tensions over the organization’s direction and priorities.

OpenAI will not comment on whether the divergence of research directions led to a split, but said that when Amodei left, it had made internal changes to more closely integrate its research and security work.

Microsoft has obtained exclusive rights to use OpenAI research results Commitment of 1 billion U.S. dollars Support the group, most of which support its computationally intensive deep learning systems in technical form, including GPT-3. Earlier this week, Microsoft stated that it has embedded the language system into some of its software creation tools so that people without coding skills can create their own applications.

Anxious for rapid commercialization GPT-3 It is in sharp contrast with OpenAI’s handling of the early technology version developed in 2019. The organization initially stated that it would not release technical details about this breakthrough due to concerns that its powerful language system might be abused, although it later changed its direction.

In order to avoid commercial interference, Anthorpic has registered as a public welfare company or “B corp” and has adopted special governance arrangements to protect its mission of “responsibly developing and maintaining advanced artificial intelligence for the benefit of mankind”. These include the establishment of a long-term welfare committee composed of people who have no relationship with the company or its supporters, and they will have the final decision on matters including the composition of the board of directors.

Anthropic stated that its work will focus on “large-scale artificial intelligence models,” including making systems easier to interpret and “building methods that more closely integrate human feedback into the development and deployment of these systems.”

Daily newsletter

#techFT brings you news, comments and analysis on big companies, technologies and issues, which are made up of experts from all over the world. click here Get #techFT in your inbox.